You may use the information and images contained in this document for non-commercial, personal, or educational purposes only, provided that you (1) do not modify such information and (2) include proper citation. If material is used for other purposes, you must obtain written permission from the author(s) to use the copyrighted material prior to its use.

Low-cost mapping of intertidal reefs using side-scanning sonar and UAV systems

Mustafa Mokrech, George Guillen and James YokleyEnvironmental Institute of Houston and College of Science and Engineering, University of Houston-Clear Lake

Houston, Texas 77058

Background and ObjectivesOyster reefs provide important environmental services includingwater filtration and purification, protection of seagrass beds andsaltmarshes from wave action, forage for some invertebrates andfinfish, wading birds, and habitat for numerous marine organismsincluding commercially and recreationally important finfish.Mapping the physical extent and conditions of these reefs providesscientists and managers with data on the current status of oysterpopulations and hard bottom habitat within an estuary as well asinformation needed for ongoing oyster conservation and restorationefforts. Mapping intertidal shallow reefs are usually impeded by theinability to use traditional subtidal side scan sonar and surveymethods. This ongoing research project aims to apply side-scanningsonar system (subtidal water depth 0.9-1.5 meters) combined withdrone (also known as Unmanned Aerial Vehicle, UAV) photography(for shallow water where water depth < 0.9 meters) together withdigital image processing techniques to map selected intertidal oysterreefs and shell bottom. Conclusions and Further Analysis

1. Drone photography can be used in mapping the extent of intertidal reefs during extreme low water levels during full/partial exposure of reefs.2. The spatial distribution and shapes of reefs must be considered during the planning process of UAV flights as sufficient key points must exist in each aerial image in order to conduct successful digital image processing of the collected data.3. The side-scanning sonar system is useful for mapping the extent of shallow reefs. The sonar images may overlap with the images collected by drones.4. Analyzing the composition of oyster reefs is a challenging process in digital image processing as submerged and exposed reefs produce different signatures. Further analysis in this regard combined with visual image interpretation in still underway.5. The logistics for acquiring aerial images using drones can be challenging over shallow waters. Using the right operational platform (e.g. air boat) for water transport is essential6. Although acquiring FAA permits to operate UAVs is time consuming, the return value of this technology can be significant.

AcknowledgmentsWe would like to thank the Gulf Coast Prairies LLC for funding thisproject. We also appreciate all the assistance in the field from the staffand graduate students at Environmental Institute of Houston, UHCL.

Contact for further information and questionsmokrech@uhcl.edu

Methods and Initial ResultsSide-scanning sonar: Transect routes were established to capturesonar images that were partially overlapped to ensure no gaps existedin the final products as well as videos for comparison purpose. Sidescan images were captured using a Humminbird Model 1197c with aXHS 9 HDSI 180 T Transducer attached to a small draft aluminumboat or kayak (Fig 2 & 3). Data collection was conducted during calmwind and at approximately high tide time to ensure proper functioningof survey equipment in shallow water. Post processing was conductedusing SonarTRX and ArcGIS (Example is shown in Fig 4).

Fig 1. Three sites are considered: Site1 is already surveyed; Site 2 is scheduled to be surveyed; and Site 3 is considered for survey subject to FAA approval.

Fig 6. A flying acquisition plan in the Bastrop Bay shows the flying path that includes the launching and navigation part, surveying path along the reef and the navigation part to landing.

Fig 2. Operating the side-scanning sonar system with a transducer installed in the front of a small vessel.

Fig 7. Densified point cloud: green dots indicate where the UAVis, blue dots where the UAV should be at the time of the image capture, and red dots indicate discarded data.

Fig 3. Operating the side-scanning sonar system in shallow water from a kayak.

Fig 4. Mosaic of the sonar images collected using the side-scanning sonar system. The processing is conducted using SonarTRX software.

Fig 8. 2D orthoimage of the data acquired using UAV at 200ft altitude: the spatial resolution is 2.2.

Fig 9. Closer display of part of the orthoimage showing the levelof details captured by the UAV camera.

Fig 5. Field data collection of the reef composition at 10m×10m site using 20 random samples.

Fig 10. Inspecting the composition of live oyster, dead oyster and shell hash within 1m×1m quadrat. 9199

Study SitesThe following criteria were considered when selecting our twoprimary sites (Fig 1) for conducting this research: 1) sufficient waterdepth to allow easy and safe navigation of a small boat equipped withside scanning sonar system to the site; 2) high potential for aerialexposure of the reef sites due to low tides or wind driven winterstorm systems; and 3) sites that are distant from airports to complywith all Federal Aviation Administration (FAA) flight restrictions. Athird site (Site 3) did not satisfy the last criterion but is currentlyunder consideration due to its high ecological importance and we arecommunicating with FAA to gain future flight approval.

Drone PhotographyThe low-cost low altitude (200–300ft) photography using a QuestUAV (Aqua) with a photogrammetric system was used to acquirehigh-resolution images for mapping intertidal reefs at a reasonablecost (Fig 1). The collected images were digitally processed usingPix4D software to produce an ortho-image using a workflow thatincludes establishing the orientation of images, key points, imagematching, and point cloud densification. The final product (Fig 8 & 9)was a 2.2 cm spatial resolution image where the reef extent can beclearly identified and mapped. The investigation of reef composition isunderway using both visual interpretation and digital image processingthat includes hard and soft classification using the ArcGIS and IDRISIplatforms. Training data collected using 20 random quadrats (1m×1m)within 10m×10m area (Fig 5 & 10) was used. In addition, thecomposition within 1m×1m quadrats at additional sites in differentlocations on the reefs were used for validation purposes.